Search

Abstract Reliable and ultra-fast DNA and RNA sequencing have been achieved with the emergence of high-throughput sequencing technology. When combining the results of DNA and RNA sequencing for tumor cells of cancer patients, neoantigens that potentially stimulate the immune response of either CD4+ or CD8+ T cells can be identified. However, due to the abundance of somatic mutations and the high polymorphic nature of human leukocyte antigen (HLA) it is challenging to accurately predict the neoantigens. Moreover, comparing to HLA-I presented peptides, the HLA-II presented peptides are more variable in length, making the prediction of HLA-II loaded neoantigens even harder. A number of computational approaches have been proposed to address this issue but none of them considers the DNA origin of the neoantigens from the perspective of 3D genome. Here we investigate the DNA origins of the immune-positive and non-negative HLA-II neoantigens in the context of 3D genome and discovered that the chromatin 3D architecture plays an important role in more effective HLA-II neoantigen prediction. We believe that the 3D genome information will help to increase the precision of HLA-II neoantigen discovery and eventually benefit precision and personalized medicine in cancer immunotherapy.

Study of flow behavior within rod bundles has been an active topic. Surface modification technologies are important parts of the design of the fourth generation reactor, which can increase the strength of the secondary flow within the rod bundle lattices. Quasi-periodic large-scale vortex structure (QLVS) is introduced by arranging micro ribs on the surface of rod bundles, which enhanced the scale of the secondary flow between the rod bundle lattices. Using computational fluid dynamics (CFD) and water experiments, the flow field distribution and drag coefficient of the rod-bundle lattices are studied. The secondary flow between the micro-ribbed rod-bundle lattice is significantly enhanced compared to the standard rod-bundle lattice. The numerical simulation results agree well with the experimental results.

Distributed machine learning, as a popular computing architecture for artificial intelligence, still faces challenges of slow model training and poor data performance transmission.Traditional network modalities were un able to meet the communication needs of distributed machine learning scenarios, hindering the improvement of model training performance.New network modalities and operation logic for distributed machine learning scenarios using multimodal network technology were proposed.This approach was designed based on application characteristics and provides implications for the use of multimodal network technology in various industries.

Introduction Osteoporosis is a major cause of fractures and even life-threatening fractures in the elderly. Mind-body exercise is a beneficial intervention to improve flexibility, control body balance and reduce pain. We aimed to evaluate the effects of physical and mental exercise on osteoporosis in the elderly. Methods Randomized controlled trials (RCTs) focusing on mind-body exercises for osteoporosis were included. Web of Science, PubMed, Science Direct, Medline, Cochrane Library, China National Knowledge Infrastructure (CNKI), and Wanfang were searched from inception to January 2023. Outcomes included bone mineral density (BMD), bone mineral content (BMC), body balance (BB), pain, indicators of bone metabolism (BMI), lower extremity function, fearing level, and quality of life (QOL). The quality of study reporting was rated by 2 reviewers independently, and Review Manager software (version 5.3) was used for meta-analysis. Results Thirty-nine trials with 2325 participants were included. The pooled results showed that mind-body exercises have encouraging effect on elderly people with osteoporosis, especially in aspects of BMD, BMC, QOL, improving the function of lower extremity, reducing pain and fearing level. While, dance and eight-section brocade could not improve the quality of life,or dance and eight-section brocade have no effect on BMD. Conclusions Mind-body exercises may have potential efficacy for osteoporosis in the elderly. However, due to the poor methodological quality of the included trials, more clinical trials with precise methodological design and rigorous reporting are needed.

Abstract This study validates the ultra-high-degree gravity field models in terms of the internal error estimate and the external precision. Internal error estimate is evaluated by geoid error degree variance and cumulative geoid height errors. The evaluation of the external precision is carried out using observed ground gravity data sets in Qinghai-Tibet Plateau and Sichuan Basin of mainland China. The results show that the geoid degree error is at the millimeter level, and the accumulated geoid error is at the centimeter level, and SGG-UGM-2 has the highest precision in terms of geoid errors. However, in terms of gravity anomaly, the GECO model has the highest precision of 37.080 mGal in the Qinghai-Tibet Plateau, and after terrain correction, its precision can reach 28.907 mGal, an improvement of 22%. In the Sichuan Basin, EGM2008 performs best with a precision of 7.202 mGal; the precision of EGM2008 becomes 6.648 mGal after terrain correction. These results mean that the terrain correction must be considered in the area where topography varies largely, while when the terrain is relatively flat, the effect of terrain on gravity can be ignored.

The dimension reduction (DR) technique plays an important role in hyperspectral image (HSI) processing. Among various DR methods, superpixel-based approaches offer flexibility in capturing spectral–spatial information and have shown great potential in HSI tasks. The superpixel-based methods divide the samples into groups and apply the DR technique to the small groups. Nevertheless, we find these methods would increase the intra-class disparity by neglecting the fact the samples from the same class may reside on different superpixels, resulting in performance decay. To address this problem, a novel unsupervised DR named the Collaborative superpixelwise Auto-Encoder (ColAE) is proposed in this paper. The ColAE begins by segmenting the HSI into different homogeneous regions using a superpixel-based method. Then, a set of Auto-Encoders (AEs) is applied to the samples within each superpixel. To reduce the intra-class disparity, a manifold loss is introduced to restrict the samples from the same class, even if located in different superpixels, to have similar representations in the code space. In this way, the compact and discriminative spectral–spatial feature is obtained. Experimental results on three HSI data sets demonstrate the promising performance of ColAE compared to existing state-of-the-art methods.

Abstract Background High-throughput sequencing technology has yielded reliable and ultra-fast sequencing for DNA and RNA. For tumor cells of cancer patients, when combining the results of DNA and RNA sequencing, one can identify potential neoantigens that stimulate the immune response of the T cell. However, when the somatic mutations are abundant, it is computationally challenging to efficiently prioritize the identified neoantigen candidates according to their ability of activating the T cell immuno-response. Methods Numerous prioritization or prediction approaches have been proposed to address this issue but none of them considers the original DNA loci of the neoantigens from the perspective of 3D genome. Based on our previous discoveries, we propose to investigate the distribution of neoantigens with different immunogenicity abilities in 3D genome and propose to adopt this important information into neoantigen prediction. Results We retrospect the DNA origins of the immuno-positive and immuno-negative neoantigens in the context of 3D genome and discovered that DNA loci of the immuno-positive neoantigens and immuno-negative neoantigens have very different distribution pattern. Specifically, comparing to the background 3D genome, DNA loci of the immuno-positive neoantigens tend to locate at specific regions in the 3D genome. We thus used this information into neoantigen prediction and demonstrated the effectiveness of this approach. Conclusion We believe that the 3D genome information will help to increase the precision of neoantigen prioritization and discovery and eventually benefit precision and personalized medicine in cancer immunotherapy.

To improve the network resources utilization and the quality of service, the provision of an adaptive and customizable network service is deemed a feasible approach. In this paper, based on the quality of service (QoS)-aware traffic classification and real-time network status, an adaptive update mechanism is presented to change the traditional rigid update techniques in software-defined networking. The developed update mechanism aims at abstracting the common update mechanism into update operations and calculates the update sequence on the operation granularity. The mechanism has three work modes, and each mode has a corresponding algorithm. It can adjust the work modes adaptively based on the network condition and the flow QoS requirements to improve the performance. The experimental results demonstrate that the three work modes can achieve optimal performance in ternary content addressable memory (TCAM) overhead reduction, delay, and bandwidth consumption, respectively. For example, when the tri-fusion work mode is leveraged, it provides at least an 85% reduction of the additional TCAM overhead and improves by at least 9%, 65%, and 82% compared to other work modes and the compared algorithms.

Many scientific workflow applications are moving to clouds. A scientific workflow is a complicated scientific computing task consisting of many sub-tasks, and each sub-task execution can generate the intermediate data used for the successor sub-task execution. The correct execution of scientific workflows depends on the security of the intermediate data, which is transmitted frequently between virtual machines during the process of the workflow execution. In multi-tenant clouds, the intermediate data contains three attributes: availability, confidentiality and integrity. If the intermediate data is lost, stolen, or tampered with by malicious tenants, the intermediate data’s attribute will be damaged, causing workflow interruption, the leakage of secret information or incorrect workflow results. For these problems, we propose ACISO scheme to secure the intermediate data by improving its availability, confidentiality, and integrity. In the scheme, availability, confidentiality and integrity strategy pools are constructed by various erasure codes, encryption algorithms and hash functions, respectively. Then we present a security strategy optimal allocation model named SSOA, which aims to maximize the overall intermediate data security strength while meeting the constraints of the workflow makespan and storage overhead. Normally, a scientific workflow contains a large number of the intermediate data, so solving this model is NP hard. Therefore, we propose a heuristic solution to solve SSOA. The simulation results show that ACISO can effectively improve the availability, confidentiality, and integrity of the intermediate data of the scientific workflows.

Software defined networking (SDN) realizes the full decoupling between control plane and data plane, and it could effectively simplify network management and promote network innovation. In the large SDN network, there are the single point of failure and the lack of control resources for one centralized SDN controller, so several researchers have presented deploying multi-controller to improve the reliability and scalability of the network. However, unreasonable deployment of multiple controllers may cause the unbalanced distribution of controller loads, and it is harmful to the network stability. In order to address the above problems, this paper proposes bidirectional matching strategy (BMS) and designs its corresponding algorithm to implement multi-controller deployment. First, through collecting the information of hops, delay, and traffic in the network, we build the matching lists of switches and controllers, respectively. Then, according to the specific sequencing of elements in the matching lists, we design a mutual selection strategy that the switch and controller will select the optimal elements from its list to implement matching in turns until realizing the construction of the distributed network. Meanwhile, we implement the strict mathematical derivation to demonstrate the feasibility of BMS. Finally, simulation shows that BMS can achieve the better multi-controller deployment when compared with the existing algorithms.

Software-defined networking (SDN) is a novel network paradigm that enables flexible management for networks. As the network size increases, the single centralized controller cannot meet the increasing demand for flow processing. Thus, the promising solution for SDN with large-scale networks is the multi-controller. In this paper, we present a compressive survey for multi-controller research in SDN. First, we introduce the overview of multi-controller, including the origin of multi-controller and its challenges. Then, we classify multi-controller research into four aspects (scalability, consistency, reliability, and load balancing) depending on the process of implementing the multi-controller. Finally, we propose some relevant research issues to deal with in the future and conclude the multi-controller research.

This paper proposes DROM, a deep reinforcement learning mechanism for Software-Defined Networks (SDN) to achieve a universal and customizable routing optimization. DROM simplifies the network operation and maintenance by improving the network performance, such as delay and throughput, with a black-box optimization in continuous time. We evaluate the DROM with experiments. The experimental results show that DROM has the good convergence and effectiveness and provides better routing configurations than existing solutions to improve the network performance, such as reducing the delay and improving the throughput.

The plasma particle and power exhaust to the divertor surface drives both particle and power recycling at the surface, which in return constrains the plasma density and temperature at the target and their profile further upstream. Both particle and power exhaust fluxes are mediated by the plasma sheath next to the divertor surface. In particular, the Bohm criterion constrains the ion exit flow speed, which enters directly into the particle flux and the kinetic flow energy component of the ion power flux, and indirectly into the electron power flux through the sheath potential drop. Here we give an overview on how the Bohm speed is set in a general plasma and how it enters power exhaust and power recycling at the divertor surface, and the implication on the correct implementation of sheath boundary conditions in numerical codes. The cases of ideal and non-ideal Bohm speed are distinguished as a result of the physics discussion. Keywords: Bohm criterion, Ideal Bohm speed, Non-ideal Bohm speed, Plasma particle exhaust, Plasma power exhaust, Power recycling, Ion reflection